n.tolstov/ai-core-api-client
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

use dedicated span per conversation_post for token usage tracking
Some checks failed
ai_core_api_client Python package / build (3.10) (push) Failing after 6s
Details
ai_core_api_client Python package / build (3.11) (push) Failing after 4s
Details
ai_core_api_client Python package / build (3.12) (push) Failing after 4s
Details
ai_core_api_client Python package / build (3.13) (push) Failing after 4s
Details
ai_core_api_client Python package / build (3.9) (push) Failing after 4s
Details

Browse Source
This commit is contained in:
NikitolProject 2026-03-26 13:31:32 +03:00
parent 395bbdbaf8
commit 0099bc4bbd
1 changed files with 10 additions and 11 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

7
ai_core_api_client/api/ai_api.py
View File

@ -27,15 +27,14 @@ from ai_core_api_client.rest import RESTResponseType
def _record_usage_to_span(response) -> None:
"""Record token usage from ai-core-api response as OTel span attributes."""
"""Record token usage from ai-core-api response as a dedicated OTel span."""
try:
usage = getattr(response, 'usage', None)
if usage is None:
return
from opentelemetry import trace
span = trace.get_current_span()
if not span.is_recording():
return
tracer = trace.get_tracer("ai-core-api-client")
with tracer.start_as_current_span("ai_core_api_usage") as span:
model = getattr(usage, 'model', None)
if model:
span.set_attribute("gen_ai.usage.model", model)